Hi chiang, What I meant was TCP acknowledgement. Of course for whatever reason >
TCP acks don't help. In fact, no ack can ever help. Imagine that the client stores the delta on hard disk thus that it cannot be lost. Then it sends it to the server via TCP. In the very moment when the server sends the TCP ack, the client crashes. When the client restarts it cannot know whether its data has arrived at the server or not. QED. This falls back to the famous "two-army problem" or "two-general problem". In an asynchronous system (which the internet is) no two parties can ever guarantee reaching an agreement when there are message losses and no time bounds. It is well documented in text books about distributed systems, thus I think every ACK based solution cannot be appropriate. It can become arbitrarily unlikely that something goes wrong, but it is not perfect. Cheers Torben the server can behave strangely and does not update the delta change > to the document. But I would have thought that the server WILL apply > the change unless it crashes or something. Thus the latest delta > change would not have taken effect, and the client will need to track > back to its previous version of the document. I think the problem lies > in the FedOne console client not having OT implemented. This means > that the client will be unable to track back to its previous version > of the document, I think. As it is, with regards to the FedOne console > client, it will retrieve the whole document or the full history (less > the latest change) from the server (and I mean the sandbox server > here, not the FedOne server). > > I agree with you that the FedOne console client is not a reference > implementation for wave client, and there are still much to do between > client and server... > > cheers, > > Chiang > > -- > You received this message because you are subscribed to the Google Groups > "Wave Protocol" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]<wave-protocol%[email protected]> > . > For more options, visit this group at > http://groups.google.com/group/wave-protocol?hl=en. > > > > -- --------------------------- Prof. Torben Weis Universitaet Duisburg-Essen [email protected]--
You received this message because you are subscribed to the Google Groups "Wave Protocol" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to [email protected].
For more options, visit this group at http://groups.google.com/group/wave-protocol?hl=en.
